the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 18 Jul 2025
The dependence of albedo on different factors for refreezing melt ponds in the Arctic
Abstract. Sea ice plays an important role in the heat transfer into the Arctic Ocean whereas the presence of melt ponds on sea ice complicates the scenario. In this study, we report a series of observations conducted in the central Arctic during 2012–2020 to investigate the optical and physical properties of refreezing melt ponds. From early August to early September, the albedo of ponds in the Pacific sector increases by 0.0036 d-1, which is attributed to the changes on surface state. Based on the typical albedo, the types of melt ponds were categorized as water pond (0.14), water-ice pond (0.20), ice pond (0.25), ice-snow pond (0.39) and snow pond (0.74). Further analysis reveals the capacity of different ratios of spectral albedo on the distinction between snow ponds and unponded ice. In addition, the total albedo of ice ponds decreases with rising pond depth, and the increasing of ice lid thickness reduces the albedo while increases that of ice-snow ponds. Based on the observations, we modified a two-stream radiative transfer model, reducing its remaining error from observation by an order of magnitude. The simulation indicates ice lid thickness as the most important determining factor in the total albedo during the freezing process.
- Preprint
(2614 KB) - Metadata XML
-
Supplement
(79008 KB) - BibTeX
- EndNote
Status: open (until 23 Oct 2025)
- RC1: 'Comment on egusphere-2025-2011', Anonymous Referee #1, 30 Sep 2025 reply
-
RC2: 'Reply on RC2', Anonymous Referee #2, 03 Oct 2025
reply
General comments
The manuscript from Zhu et al. identifies the role of ice-lead thickness, melt-pond depth, and substrate ice thickness in total and spectral albedo values of melt ponds. Melt ponds have a large impact on sea ice albedo during the summer and fall. Moreover, their characteristics also differ during the refreezing season (August-September) compared to the fast-melting period (June-July). Therefore, this study addresses relevant aspects of Arctic sea ice life cycle. Results from the observations of total and spectral albedo, along with the radiative transfer simulations, are well described with comprehensive step-by-step explanations. However, some effort is required to introduce the radiative transfer model, to avoid copying and pasting equations and text from Lu et al. (2016), and instead to highlight what was adapted in the model for this specific analysis. The discussion section will also benefit from a comparison between the observations and satellite measurements (see specific comments). The summary should be expanded to a conclusion that identifies the strengths and limitations of this study and outlines the next steps to advance the analysis.
Specific comments
Line 21: “and thus control the radiative forcing in the Arctic Ocean and the world (Hudson, 2011)”. Would it be possible to rephrase the sentence for more accuracy, avoiding “the world” which sounds simplistic in the context of the paper?
Line 65: “short-term and long-term ice stations” could these stations be identified in Table 1 or in Figure 1? Did you notice significant discrepancies in total albedo or spectral albedo due to the sampling duration?
Line 92: lambda in the equations should be defined.
Section 2.2 this section has to be significantly reworked to specify what differs from than Lu et al. (2016) and to clearly identify the novel elements of the present analysis.
Figure 3: it should be acknowledged in the figure legend that it has been adapted from Lu et al. (2016).
Lines 112: the equations are exactly the same than in Lu et al. (2016) along with the description until line 117. It should be acknowledged. It should also be better to highlight how adding the ice lid is impacting the equations and what is different in the calculations compared to previous studies.
Line 119: in which equations/calculation R1 and R2 are used in the analysis?
Line 167: “As a result, the calibration reduces the median deviation of CNR4 measurements from 0.2 to 0.06” is this reduction applied to all types of melt ponds? Maybe add some values in Fig. 5c and 5d to highlight the impact of the calibration.
Figure 6b: what does the colorbar represent? What about the numbers and uncertainty/standard deviation in red? MP96, IA94, etc. are not defined yet.
Line 278: “which is consistent with observation in Malinka et al. (2016)” Can you elaborate more on the agreement between your observations and previous studies?
Line 288: Can you better introduce α412/α667 as no results are presented in section 3.2 about α412/α667 and it is only in section 4 that references are made to this ratio.
Line 290: “that some uncertainty remains in this result” Can you be more accurate about the uncertainties? Can the uncertainty be quantifiable?
Line 299: “a correlation coefficient of 0.12 is found between” how is it calculated? Using values from figure 7 and figure 9?
Line 333: “a radiative transfer model” add “described in Section 2.2.
Line 394: “except for the influence of temperature and radiation which is discussed in section 3.2”, the influence should be reminded to complement the discussion.
Section 4 discussions:
Wavelengths corresponding to MODIS bands are selected to identify the limits of the ratio used to identify snow covered pond. The analysis would be stronger if some in situ observations were compared with collocated MODIS measurements to assess how effectively the satellite performs and under which conditions MODIS is too limited. If a case study cannot be conducted, some references to MODIS pond identification should be cited and compared with the present study.
Two melt pond retrieval algorithms developed for satellite data are applied to the observations from the present study. Although references for both algorithms are provided, it would be helpful to introduce their underlying principles and to specify if any adaptations made to use them with in situ measurements. This could be included in the Supplementary Information to complement the study’s methodology. Again, a case study comparing observations and Sentinel-2 data would be a valuable addition to extend the analysis to satellite observations.
Section 5 Summary:
The summary should be expanded into a forward-looking conclusion that clearly states the study’s limitations and outlines concrete next steps. What is still required to improve the analysis and reduce the uncertainties? What would be necessary if these observations were conducted again (e.g., meteorological data and snow depth, etc.)? How will the parametrization of the radiative transfer model be used? Is there a potential study to improve the identification of snow-covered ponds from satellite? What kind of refreezing melt pond studies does the community need to advance understanding of sea ice albedo?
Technical comments
Line 92: the equations should be numbered.
Line 317: More consistency should be applied for defining the acronyms to avoid confusion. Some acronyms are defined in Figure 10 legend and then used in the text without explanation: line 312 “the fitting result is close to that of ML96 and SP07”. Etc.
Line 416: the opposite is also observed, PCA algorithm and LinearPolar Algorithm are defined in the text line 418 but not in the Figure 13 legend.
Figure 11: unit of H should be defined.
Figure 13: use color-coded makers as in previous figures to help readers distinguish between the different types of met ponds.
References: Rösel and Kaleschke, 2011 is missing in the bibliography.
There are two references for Wang et al. (2020). They should be labeled a and b to avoid confusion, or differently to distinguish the different authors.
Citation: https://doi.org/10.5194/egusphere-2025-2011-RC2
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 1,092 | 31 | 33 | 1,156 | 30 | 53 | 52 |
- HTML: 1,092
- PDF: 31
- XML: 33
- Total: 1,156
- Supplement: 30
- BibTeX: 53
- EndNote: 52
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Abstract:
The statement of the research objective should be clearly articulated. The goal of investigating the optical and physical properties of refreezing ponds must be explicitly stated. What exactly is the spectral range? What are the surface characteristics of the ice-cover? This is not entirely clear in its current form. The abstract should clearly communicate that the primary objective is to generate a dataset focused on albedo-based classification of surface states, with particular emphasis on refrozen melt ponds (which remain under-documented compared to seasonal albedo variations). The exact goal should be distinctly stated—what and why. If there are multiple objectives, they could be presented in order of priority. For example, the development of the methodology could be considered one of the objectives, along with the spatial and temporal variation of albedo and the relationship between physical and optical properties.
General comments: he albedo categorization is unclear. What is the basis for the threshold? Since the distinction of 'classes' or types is vital, it may benefit from a brief explanation of the rationale behind this threshold. Why does refreezing specifically matter, and how does your classification fill a particular gap?
Introduction:
In-situ Data:
Methodology:
Results: